Parallel metaheuristic algorithm for route planning using CUDA
This research will be focusing on developing a Parallel Metaheuristic Algorithm for Route Planning using CUDA to improve the efficiency and performance of route planning. The increasing demand for more efficient route planning approaches, which includes several factors such as cost savings, timely d...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project / Dissertation / Thesis |
| Published: |
2025
|
| Subjects: | |
| Online Access: | http://eprints.utar.edu.my/6157/1/fyp_CS_2025_LDJJ.pdf http://eprints.utar.edu.my/6157/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This research will be focusing on developing a Parallel Metaheuristic Algorithm for Route Planning using CUDA to improve the efficiency and performance of route planning. The increasing demand for more efficient route planning approaches, which includes several factors such as cost savings, timely deliveries and reduced carbon emissions, has led to a surge in demand for more advanced route planning algorithms in search of more efficient solutions.
The problem that this research will be tackling is the Travelling Salesman Problem (TSP), which is a specific type of route planning problem where the main objective of it is to find out the optimal set of routes for a given number of vehicles to transport goods to a defined set of destinations. TSPs are known to be NP-hard problems[1] where an increase in the number of vehicles and destinations will significantly increase the computational time required to obtain an optimal solution. Existing works that utilized metaheuristic algorithms have shown their flexibility in solving multiple TSP variants and their capabilities in obtaining near-optimal solutions within a reasonable amount of time. However, due to the limitations of CPUs in terms of parallelization, these algorithms do not perform well as they are highly iterative.
The proposed approach will be utilizing the Compute Unified Device Architecture (CUDA) to enhance the performance and efficiency of metaheuristic algorithms in finding optimal solutions for the TSP by leveraging the parallel processing capabilities of Nvidia Graphics Processing Units (GPUs). This research aims to significantly speed up solution searching for the TSP by using GPUs compared to CPUs. Besides, this research strives to provide a foundation for future research on parallel metaheuristic algorithms, and to further encourage their implementations in real-world instances of route planning.
Area of Study: Massively Parallel Computing, Combinatorial Optimization
Keywords: Parallel Metaheuristic Algorithm, Travelling Salesman Problem, CUDA, GPU, Genetic Algorithm |
|---|